ELSEVIER Acta Tropica 62 (1996) 257-267

ACTA TROPICA

Female genital schistosomiasis (FGS): relationship between gynecological and histopathological findings

Gertrud Helling-Giese a.b, Aimee Sjaastad o, Gabriele Poggensee d, Eyrun Floerecke Kjetland c, Joachim Richter d, Lester Chitsulo e,

Newton Kumwenda e, Paul Racz b Borghild Roald c Svein Gunnar Gundersen ~, Ingela Krantz f, Hermann Feldmeier d,s,,

a Department of Gynecology and Obstetrics, Frauenklinik Finkenau, Hamburg, Germany b K~rber AIDS Research Laboratory, Department of Pathology,

Bernhard-Nocht-lnstitut fffr Tropenkrankheiten, Hamburg, Germany Department of Pathology and Ullevaal Center for International Medicine,

Research Forum and Department of Infectious Diseases, Ullevaal Hospital, University of Oslo, Oslo, Norway

d Institute of Tropical Medicine, Berlin, Germany Ministry of Health and Environmental Affairs, Lilongwe, Malawi

f Nordic School of Public Health, G6teborg, Sweden g Faculty of Medicine, Free University, Berlin, Germany

Abstract

Schistosomiasis of the lower female reproductive tract manifests itself in a broad spectrum of clinical features. However, clinical and histopathological findings have never been studied in a synoptic manner. Based on the assumption that any type of pathology present in the female reproductive tract is the expression of a complex pathophysiological reaction towards eggs sequestered in the genital tissues, we decided to analyze colposcopic and histopathological findings in a comprehensive manner. Thirty-three women in Malawi with urinary and genital schistosomiasis were examined parasitologically and gynecologically. A thorough colposcopic examination with photodocumentation was performed and biopsies were taken from the cervix, the vagina and/or the vulva for histological sectioning and immunohistochemistry. The predominant colposcopic findings were sandy patches on the cervical surface similar to those seen in the bladder and polypous/papillomatous tumors with irregular surface on the vaginal wall and in the vulvar area. The histopathological sections of sandy-patch-like lesions demonstrated only a small cellular reaction around S. haematobium eggs in various stages of disintegration. In contrast, in the case of polyps the histology revealed a more pronounced immunological reaction characterized by a heavy cellular infiltrate. One case of invasive squamous cell carcinoma of the cervix was diagnosed. We conclude that colposcopy is a useful tool in the detection of FGS related pathology in the lower female reproductive tract

* Corresponding author. Am Rain 7, D-21244 Buchholz, Germany. Fax + 49 4181-36943.

0001-706X/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved PII S0001-706X (96)00027-7

258 G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267

and that the synoptic assessment of surface and of corresponding histological sections helped to understand the pathophysiology of S. haematobium associated disease in genital tissue.

Keywords: Female genital schistosomiasis; Colposcopy; Histopathology; Diagnosis; Pathophysiology; Eosinophils

1. Introduction

Schistosomiasis of the lower female reproductive tract has many gynecological facets. Among other cauliflower-like lesions, ulcers, polyps, vaginal warts, granulo- matous lesions and general hyperplasia of the genital epithelium have been described as typical alterations of the cervix, the vagina and the vulva (Madden, 1899; Badawy, 1962; Attili et al., 1983; Friedberg et al., 1991). However, many of the observations are based on case-reports not allowing definite conclusions on the specificity of the findings (Sharma et al., 1970; LtRtges and Koransky, 1985; Mawad et al., 1992; Goldsmith et al., 1993). A systematic classification of colposcopic findings in girls and women with female genital schistosomiasis (FGS) has never been attempted so far.

In contrast, several thorough histopathological studies have been carried out in which characteristic tissue reactions of frequently affected topographic sites of the female genital tract have been identified (Nosny, 1963; Berry, 1966; Williams, 1967; Youssef et al., 1970; Wright et al., 1982; Friedberg et al., 1991). In these studies, though, the gynecological findings are lacking or are described rather vaguely, leaving the tropical doctor with a limited understanding of the relationship between histopathology and macroscopicaUy detectable alterations. Based on the assumption that any type of pathology present in the female genital tract is the expression of complex pathophysiological reactions towards eggs sequestered in the epithelium or subepithelial tissue, we decided to analyze colposcopic and histopathological findings in a synoptic manner. To this end 33 women from rural Malawi in whom FGS was confirmed parasitologically, were examined gynecologically using a standardized protocol. The results were analyzed in relation to histopathological findings in biopsies taken from the same patients.

2. Materials and methods

2.1. Study design and study population

The study design, the study population and the parasitological investigations performed are described in an accompanying paper (Kjetland et al., 1996).

For differential diagnosis of lesions we performed Gram stains for detection of Neisseria gonorrhoeae. Serum was tested for the presence of antibodies against Treponema pallidum using the VDRL test in suspected cases. HPV infection was diagnosed when there was histological presence of koilocytosis.

G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267 259

2.2. Gynecological examination

After inspection of the vulvar and perianal area, of the vaginal mucosa and of the cervix with the naked eye, a thorough colposcopic examination was performed. All pathological findings were documented using a built-in camera..Lesions were recorded in accordance with a modified WHO protocol (WHO, 1995). This protocol included a description of the epithelium, blood vessels, demarcation zone and slough and accounted for the cervix, the vagina, vulva and perineum/perianal area separately.

In each patient a cervical biopsy was performed. When cervical pathology was present, the biopsy was taken from the lesion, otherwise the site for a biopsy was chosen at random.

For ethical reasons, biopsies of the vagina or the vulva were only performed if lesions were present. A patient was considered to have FGS if S. haematobium eggs were detected in fresh crushed biopsy material or by histological sectioning of formalin-fixed tissue (Kjetland et al., 1996).

2.3. Histology and immunohistochemistry

The biopsies were stored in formaldehyde and shipped to laboratories in Germany and Norway. Microscopic examination was done on 5 ~tm thick hematoxylin-and- eosin-stained sections from the formalin-fixed, paraffin-embedded biopsies. The histopathological diagnosis of female schistosomiasis was always based on the detec- tion of eggs in histological sections.

Care was especially taken to register the presence of schistosome eggs and their localization. Eggs were classified as viable or non-viable. Ova were considered to be viable at the time the biopsy was taken when mature miracidia inside intact shells were detectable. Mature miracidia were recognized by their intact germinal cells (Wright et al., 1982). Eggs or shell fragments were otherwise classified as non-viable.

The presence or absence of epithelial hyperplasia and/or dysplasia/atypia was recorded in addition to stromal changes like grade and type of inflammatory reaction, fibrosis and vessel proliferation.

In the presence of koilocytosis (perinuclear halo of keratinocytes), immuno- histochemistry was done with antibody against human papillomavirus. Immunohistochemistry with antibody against eosinophilic cationic protein (ECP) present in the granules of eosinophilic granulocytes was also performed, as previously described (Poggensee et al., 1996). The number of ECP-positive cells in the tissue reaction was graded semiquantitatively using a scale from 0 to 3 points.

2.4. Statistical analysis

Statistical analysis was performed using the EPI INF0 5.0 Package. Odds ratio and chi-square values with Yates's correction were calculated. Where expected values were below 5, Fisher's two-sided Exact Test was used to compare the groups.

260 G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267

3. Results

3.1. Colposcopic findings

In 33 women eggs were detected in genital tissue and in the urine (FGS-positive), in 21 ova were found only in urine (FGS-negative). In these patients a characteristic pattern of genital abnormalities was observed (Table 1).

Sandy patches similar to those seen in the bladder of patients with urinary schistosomiasis occurred in 9 FGS cases, but in none of the 9 women without FGS (P=0.008).

Polypous/papillomatous tumors with irregular surface were observed in 20 out of 33 FGS eases compared to 13 out of 21 women without eggs in genital tissue (difference not significant). All other surface alterations detected by colposcopy including erosion, edema, leukoplakia, inflammation, swelling, ulceration, erythema and ectopia did not correlate significantly with the presence of eggs in genital tissue either.

In FGS patients polypous and papillomatous tumors occurred significantly more often in the vulva and vaginal wall than on the cervix (P=0.01). Sandy patches were found on the vaginal wall in only 1 case; this patient had no alterations on the cervix. In all remaining 8 cases this type of lesion was present only on the cervical surface. In only 2 women were sandy patches and polypous/papillomatous tumors associated.

3.2. Histological findings

In two thirds of the histological sections only non-viable ova were seen. Viable and non-viable ova were seen in the same biopsy in 4 cases. In the remaining cases, the eggs were exclusively viable. In only 1 case were eggs demonstrated inside the epithelial layer; in all other cases eggs were seen in the stroma, respecting the epithelial basal layer. Eggs were single or grouped in clusters.

In 2 biopsies epithelial dysplasia, CIN I and CIN II respectively, were recorded.

Table 1 Frequency of defined colposcopic findings in women with or without FGS

Colposcopic finding FGS positive* (n=33) FGS negative (n=21) Significance of difference

Sandy patches 9 Papillomatous/ 20 13 polypous tumor Swelling 3 - n.s. Inflammation 5 - n.s. Leukoplakia 6 - n.s.

P =O.O08 n . s

* For case definition, see Section 2. Some patients had multiple lesions.

G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267 261

In addition, an invasive squamous cell carcinoma was diagnosed in a biopsy with several viable and non-viable schistosome eggs.

Two distinct patterns of tissue reactions were observed. However, these patterns were seen along a spectrum of changes.

Type A: A strong inflammatory reaction characterized by a diffuse infiltration of plasma cells, lymphocytes, eosinophilic granulocytes and macrophages around sites of egg deposition, mainly around viable eggs. This reaction occurred in approxi- mately half of the biopsies. In one case massive necrosis with eosinophilic granulo- cytes was observed (Fig. 1; Table 2).

Type B: A fibrous connective tissue reaction with a minimal cellular infiltrate, best described as scar tissue. Not surprisingly this pattern was observed in the majority of cases around non-viable eggs or calcified shell fragments (Table 2).

Fig. 1. Strong cellular infiltrate around viable and non-viable eggs, vaginal wall.

262 G. Helling-Giese et aL/Acta Tropica 62 (1996) 257-267

Table 2 Classification of histopathological findings according to the presence of viable and non viable ova in

histological sections

Findings Positive findings (%)

Viable ova Non-viable ova

ECP + + 100 5 Presence of 71 33 eosinophils Fibrosis 0 50 Inflammation 100 11 Endothelial 100 22 swelling

3.3. Immunohistochemistry

The immunohistochemical detection of ECP-positive cells was clearly linked to the state of degeneration of eggs. ECP positive cells were seen in 100% of the biopsies when the morphological criteria for viable eggs were fulfilled. When ova were apparently dead, ECP reactivity was recorded in 5% of the biopsies only (Table 2).

3.4. Correlation between colposcopic and histological findings (Figs. 2 and 3)

Only sandy patches, on the cervix, seemed to be specifically related to FGS. Polypoid/papiUomatous tumors, in the vagina and vulva, were considered to be pathognomonic for FGS only in the absence of HPV infection or syphilis. The analysis of a putative relationship between gross mucosal surface alterations and the underlying histological picture was thus limited to this type of lesion.

Sandy patches on the cervix are characterized as slightly elevated lesions, with irregular borders, covered by a finely granular epithelium. At touch they had a characteristic rough, sandy surface, and on section a gritty consistency (Fig. 3). The histology usually demonstrated only a limited cellular reaction around eggs in various stages of disintegration sequestered in fibrous stromal tissue. As a rule, only a few eggs were present. Thus, this type of response belongs to pattern B as described above.

In contrast, the polypous/papillomatous tumors consisted of rounded elevations of the mucosa produced by large masses of eggs. They were located on the vaginal wall and vulva and took on various shapes such as multiple wart-like tumors, solitary, sedentary or pedunculated polyps or lesions located in groups along the rugae vaginales (Fig. 2). In the vulvar and perianal area they had a greasy appearance and were less colored than the surrounding black skin Fig. 4. The cellular reaction found in the tissue sections of polypoid material corresponded to the type A pattern of immune response characterized by a heavy infiltration of various cell types around viable-looking ova. In Table 3 the characteristic features of colposcopic and histo- pathological findings are summarized.

G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267 263

Fig. 2. Multiple papillomatous tumors of the vaginal wall.

Fig. 3. Sandy patches on cervical surface.

264 G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267

Fig. 4. Papillomatous tumor of the vulva.

Table 3 Characteristic colposcpic findings and correlating histo~athology

Abnormalities seen Topographic site Type of eggs through the found colposcope

Corresponding histology

Main features Associated features

Sandy patches Cervix Non-viable eggs, No or little cellular reaction, Fibrosis (predominantly) viable eggs epithelial hyperplasia

Polypoid/polypous Vagina and vulva Viable eggs, Strong cellular reaction, - (exclusively) non-viable eggs epithelial hyperplasia

G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267 265

4. Discussion

This paper presents an analysis of colposcopic and histopathological findings in women with FGS in a synoptic manner. Previous reports have focussed mainly on the histopathological aspects and neither correlated the histopathological and the clinical findings nor attempted to describe macroscopic alterations in a systematic manner.

The predominant colposcopic findings in women with FGS were sandy patches of the cervix and polypous/papillomatous tumors of the vagina and vulva. All other types of lesions as described in the literature (Madden, 1899; Badawy, 1962; Nosny, 1963; Wright et al., 1982; Attili et al., 1983; Friedberg et al., 1991) were in our FGS patients not significantly associated with the presence of ova in genital tissue. Sandy patches on the cervix are discrete findings and might easily be overseen when the examiner does not use a colposcope (Youssef, 1957). This could explain that in FGS sandy patches have only rarely been described before.

The topographic distribution of the lesions was quite distinct. We found sandy patches almost exclusively on the cervix whereas polyps and papillomatous tumors occurred in the vagina and vulva only. One could speculate that this phenomenon is due to intrinsic histological characteristics of the respective tissues. However, the assumption is contradicted by the observations of Friedberg et al. (1991) who in South Africa observed that schistosomal involvement of the cervix usually manifests as a warty or cauliflower-like growth, a hard nodule, a bleeding friable area or a granulomatous ulcer. In agreement with our findings, vaginal and vulvar manifesta- tions of FGS have also been found by other authors to be associated with multiple papillomas or polypoidal granulomas (Boulle and Notelovitz, 1964; Payet and Camain, 1967; Mawad et al., 1992).

The clinical observations made in our group correlated with two distinct patterns of tissue reaction: whereas sandy patches represent surface alterations of a distinct pathology characterized by a rather minimal cellular reaction to schistosome eggs, in contrast polyps of the vaginal wall and vulva indicate an intense cellular infiltration of underlying tissue.

Similar to previous observations in female genital schistosomiasis (Nosny, 1963; Berry, 1966; Williams, 1967; Youssef et al., 1970; Wright et al., 1982; Friedberg et al., 1991), we found a spectrum of histopathological reactions, the two opposing poles being a strong cellular reaction around sites of oviposition and on the other side formation of scar tissue with no apparent inflammatory cellular infiltration.

Interestingly we did not see the typical egg granuloma as described in schistosomia- sis of the urinary tract and the rectal wall (Warren and Domingo, 1970). In order to avoid bleeding at the biopsy sites, we chose minimal-size biopsy punches. Consecutively our tissue sections consisted of very small tissue samples and only rarely were we able to examine the deeper layers of underlying stroma: histologically, this may have influenced the chance to study granuloma formation in deeper layers of the stroma.

We diagnosed I case of squamous cell carcinoma; in this case neoplastic cells coexisted with FGS since we found many viable and non-viable eggs within the

266 G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267

cancerous tissue sections. Whether schistosomal lesions should be considered as a tumorigenic factor causally linked to the development of cancer of the cervix remains a matter of debate.

In several case reports (Madden, 1899; Carpenter and Lewis, 1964; Badawy, 1962; EI-Zeneiny et al., 1968; Youssef et al., 1970; Adeleye and Odjegba, 1975; E1 Tabbakh and Hamza, 1989; Vass, 1992) the possible role of S. haematobium infection as a co-factor in the development of cervical cancer has been evoked. Epithelial hyperpla- sia, basal-cell hyperactivity, and leukoplakia--abnormalities frequently observed in women with FGS--can be regarded as precancerous abnormalities induced by a chronic irritation of sensitive tissues. Tumorigenic factors secreted by the parasite itself could also play a role in the development of cervical cancer (Chen and Mott, 1989). Other authors (Williams, 1967; Bland and Gelfand, 1970; Coelho et al., 1979; Wright et al., 1982; Szela et al., 1993; Moubayed et al., 1994), however, question an etiologic role of schistosomiasis in the oncogenesis of cervical carcinoma. None of the studies is conclusive and prospective; controlled studies are urgently needed to address this issue.

In conclusion, we found a broad spectrum of genital lesions in this series of Malawian women suffering from S. haematobium infection, but only sandy patches on the cervix and vaginal tumors seemed to be specific for FGS. Macroscopic lesions correspond to distinct patterns of histopathology and certain types of lesions are confined to distinct topographic sites. The implications of these findings with respect to the reproductive and general health of the women are enormous. Particularly the possible association between FGS and cervical cancer and HIV transmission and FGS needs further consideration (Feldmeier et al., 1994).

Acknowledgements

The investigation was supported by a Director's Initiative Grant from TDR/WHO/WB. It also received support from the Kongregation der Franziskanerinnen, Salzkotten, Germany, the Research Forum and Department of Infectious Diseases, Ullevaal University Hospital and Norwegian Lions International. We are indebted to the Mangochi District Hospital staff under the directorship of Dr. V. van Oosterzee and Dr. E. van der Velde. Special thanks go to Ms. R. Maziyauya, Mr. Kalua, Mr. A. Mazombe and Mr. E. Chikzada for their kind cooperation.

References

Adeleye, J.A. and Odjegba, A. (1975) Schistosomiasis of the uterus in association with uterine fibroids. Trop. Geogr. Med. 27, 206-208.

Attili, R.V., Hira, S.K. and Dube, M.K. (1983) Schistosomal genital granulomas: a report of 10 cases. Br. J. Vener. Dis. 59, 269-272.

Badawy, A.H. (1962) Schistosomiasis of the cervix. Br. Med. J. 1,369-372.

G. Helling-Giese et al./Acta Tropica 62 (1996) 257-267 267

Berry, A. (1966) A cytopathological and histopathological study of bilharziasis of the female genital tract. J. Pathol. Bacteriol. 32, 325-338.

Bland, K.G., Gelfand, M. (1970) Effect of schistosomiasis on the cervix uteri in the African female. J. Obstet. Gynaecol. Br. Commonw. 77, 1127-1131.

Boulle, P. and Notelovitz, M. (1964) Female genital schistosomiasis. J. Obstet. Gynaecol. Br. Commonw. 2, 48.

Carpenter, C.B. and Lewis, N.G. (1964) Schistosomiasis japonica involvement in the female genital tract. J. Am. Med. Assoc. 188, 647-650.

Coelho, L.H., Carvalho, G. and Carvalho, J.M. (1979) Carcinoma in situ and invasive squamous cell carcinoma associated with schistosomiasis of the uterine cervix. A report of three cases. Acta Cytol. 23, 45-48.

Chen, M.G. and Mott, K.E. (1989) Progress in the assessment of morbidity due to S. haematobium. A review of the recent literature. Trop. Dis. Bull. 86, R1-R36.

E1 Tabbakh, G. and Hamza, M.A. (1989) Carcinoma of the uterine cervix and schistosomiasis. Int. J. Gynecol. Obstet. 29, 263-268.

EI-Zeneiny, A.H., Badawy, S.Z. and Iskander, S.G. (1968) Bilharziasis of the female genital tract. J. Egypt. Med. Assoc. 51, 543-553.

Feldmeier, H., Krantz, I. and Poggensee, G. (1994) Female genital schistosomiasis as a risk factor for HIV transmission. Int. J. Std. AIDS 5, 368-372.

Friedberg D., Berry A.V. and Schneider J. (1991 ) Schistosomiasis of the female genital tract. South Afr. Med. J. 61, 1-16.

Goldsmith, P.C., Leslie, T.A., Sams, V., Bryceson, A.D.B., Allason-Jones, E. and Dowd, P.M. (1993) Lesions of schistosomiasis mimicking warts on the vulva. Br. Med. J. 307, 556-557.

Kjetland, E.F., Poggensee, G., Helling-Giese, G., Richter, J., Sjaastad, A., Chitsulo, L., Kumwenda, N., Gundersen, S.G., Krantz, I. and Feldmeier, H. (1996) Female genital schistosomiasis due to Schistosoma haematobiurn. Clinical and parasitological findings in women in rural Malawi. Acta Trop. 62, 239-255.

Ltittges, J. and Koransky, K. (1985) Vaginalbeteiligung bei einer Bilharziose. Pathologe 6, 161-163. Madden, F.C. (1899) A case of bilharzia of the vagina. Lancet i, 1716. Mawad, N.M., Hassanein, O.M., Mahmoud, O.M. and Taylor, M.G. (1992) Schistosomal vulval granu-

loma in a 12 years old Sudanese girl. Trans. R. Soc. Trop. Med. Hyg. 86, 644. Moubayed, P., Lepere, J.-F., Mwakyoma, H. and Neuvians, D. (1994) Carcinoma of the uterine cervix

and schistosomiasis. Int. J. Gynecol. Obstet. 45, 133-139. Nosny, Y. (1963) La bilharziose grnito-urinaire. Bull. Soc. Pathol. Exot. 56, 999-1048. Payet, M. and Camain, R. (1967) Bilharziasis in French-speaking Africa. In: Mostofi (Ed.), Bilharziasis.

International Academy of Pathology Special Monograph, Springer, Berlin, pp. 133-139. Poggensee, G., Reimert, C.M., Nilsson, L.-A., Jamaly, S., Sjaastad, A., Roald, B., Kjetland, E.F., Helling-

Giese, G., Richter, J., Chitsulo, L., Kumwenda, N., Gunderson, S.G., Krantz, I. and Feldmeier, H. (1996) Diagnosis of female genital schistosomiasis by indirect disease markers. Acta Trop. 62, 269-280.

Sharma, S.D., Ziegler, O. and Trussel, R.R. (1970) A case of schistosomiasis haematobium of the cervix. Acta Cytol. 14, 305.

Szela, E., Bachicha, J., Miller, D., Till, M. and Wilson, J.B. (1993) Schistosomiasis and cervical cancer in Ghana. Int. J. Gynecol. Obstet. 42, 127-130.

Vass, A.C.R. (1992) Bilharzial granuloma of the fallopian tube. Br. J. Obstet. Gynaecol. 89, 867-869. Warren, K.S. and Domingo, E.O. (1970) Granuloma formation around S. mansoni, S. haematobium, and

S. japonicum eggs. Am. J. Trop. Med. Hyg. 19, 292-304. WHO (1995) Manual for the Standardization of Colposcopy for the Evaluation of VaginaUy Administered

Products. WHO/GPA/RID/CRD/95.10. Williams, A.O. (1967) Pathology of the uterine cervix due to S. haematobium. Am. J. Obstet. Gynecol.

98, 184. Wright, E.D., Chipangwi, J. and Hutt, M.S.R. (1982) Schistosomiasis of the female genital tract. A

histopathological study of 176 cases from Malawi. Trans. R. Soc. Trop. Med. Hyg. 76, 822-829. Youssef, A.F. (1957) 2 cases of schistosomiasis of the cervix were only detected through colposcopy.

Geburtsh. u. Frauenheilkd. 5, 445-449. Youssef, A.F., Fayad, M.M. and Shafeek, M.A. (1970) Bilharziasis of the cervix uteri. J. Obstet. Gynaecol.

Br. Commonw. 77, 847 851.